Novel fault limiting circuit breakers for dense urban grids

The £6 million UK Power Networks project, called PowerFuL-CB (Power- electronic Fault Limiting Circuit Breaker) – scheduled to start in January 2017 – aims to trial two novel fault limiting circuit breaker devices on the London 11 kV network, one being developed by ABB, for deployment in substations, the other from Applied Materials (AMAT), designed for use at distributed generation plant sites.

The decarbonisation of heat is seen as an increasingly important issue for the UK, requiring expanded use of district heating and distributed CHP.

However, fault level constraints are becoming a barrier to connecting new distributed generation in urban areas, says Ofgem, noting that, with plans for increased local generation, especially CHP, the already limited headroom in substations will be quickly exhausted.

In one scenario, for example, London could see a greater than six-fold increase in CHP connected by 2031, with around 73% of the region’s substations requiring fault-level reinforcement.

Traditional reinforcement as a connection solution could make new distributed generation financially unattractive to developers, Ofgem believes.

The last two years have seen an increase of over 2900% in the annual distributed generation application rate for London (see Figure 1). Correspondingly, the ability to offer connections at 11 kV is limited as a result of fault level constraints. As Figure 1 suggests, the alternative solution is currently to offer connections at EHV and 132 kV levels. This is much more expensive than connecting at 11 kV, and will be unaffordable for all but the largest connections, says Ofgem. The number of connection offers accepted is decreasing due to the subsequent high offer costs resulting from fault level constraints.

Ofgem notes that dense urban networks tend to have high prospective fault currents, caused by: short cable distances; large diameter cables; direct transformation (in the London case from 132 kV to 11 kV); high degree of interconnection between substations; the need, during a transformer outage, to connect the remaining transformers in parallel to share load evenly; and the load and capacity of distributed generation already connected to the network.

Connecting new distributed generation to the network further increases the local fault current level. Thus, the network’s fault rating limits the amount of distributed generation that can be connected. This is largely because most distributed generating capacity connecting in dense urban areas uses rotating machinery, eg piston-engine-based combined heat and power installations and diesel standby generators, which have the highest impact on fault current levels.

Existing solutions are either too expensive or do not meet UK network requirements, especially in dense, urban areas.

In the London area, and in other areas of the UK with high population density, there is lack of space for new equipment, and a dependence on running several transformers in parallel to provide security of supply. Unfortunately, this means that smart solutions that would work in other types of network are unsuitable. The only option available is to reinforce the network, which is generally considered too expensive for generators requesting a connection.

The result of the trials will be that two new technology applications to address fault levels will be proven on a live network. All existing fault level mitigation technologies (FLMTs) have at least one showstopper preventing their use in a setting such as the London network, says Ofgem.

PowerFuL-CB aims to prototype and validate the use of fault limiting circuit breakers (FLCBs) to enable distributed generation connections to substations that are otherwise “full” because of fault level constraints.

A FLCB is a power electronics device that blocks 100% of fault level contribution from a single transformer/bus coupler/ generator, but allows load current to flow normally before and after the fault. Like an Is-limiter, it disconnects the transformer/ bus-coupler/generator at a speed fast enough to prevent a contribution to the “break” or “make” fault level (ie, before the first current peak); but unlike an Is- limiter, it can reclose as soon as the fault has been cleared from the network.

As already mentioned, PowerFuL-CB will demonstrate two approaches that enable new generation capacity to connect to fault-level-constrained substations:

• Installation of a FLCB at a primary substation, in series with a transformer incomer or in parallel with a bus coupler. On a busbar fed from two transformers, this reduces the fault level by up to 50%, creating significant headroom for new generation connections. Unlike using an Is-limiter or running busbars “split”, it has no impact on security of supply.
• Installation of a FLCB at a generating site, in series with a generator. This prevents individual generators from causing any increase in network fault levels, which enables the connection of large amounts of distributed generation, even if the network is “full” because of fault level constraints.

In particular, the PowerFuL-CB demonstration project will entail: 

• Building a trial-ready prototype of ABB’s 2000 A FLCB. ABB has already developed its FLCB technology to TRL (technology readiness level) 4, comprising a single-phase proof-of- concept prototype that has been lab- tested at full voltage and current. During the first two years of the PowerFuL-CB project, ABB will build a three-phase, field-ready prototype suitable for trial at a London substation.

• Demonstration of ABB’s 2000 A FLCB at a primary substation. This demonstration will prove the technical performance required to release fault level headroom for new distributed generation connections, and improve understanding of the engineering and safety requirements for deploying FLCBs at substations.

• Building of a trial-ready prototype of AMAT’s 250 A FLCB. AMAT has already developed its FLCB technology to TRL 6 and it is therefore nearly ready for demonstration.

• Demonstrate AMAT’s 250 A FLCB at a generator’s premises. This project will prove the required technical performance to allow generators to connect distributed generation capacity to substations that have little or no fault level headroom, and provide improved understanding of the engineering and safety requirements for deploying FLCBs at generating/CHP facilities.

Rethinking the FLCB

A conventional circuit breaker interrupts fault current by physically separating its contacts, allowing the resulting voltage surge to form an arc between the contacts, then using various methods to extinguish the arc. A typical vacuum circuit breaker takes 40-60 ms to open its contacts, then another 10-15 ms to extinguish the arc, for a total interruption time of 50-75 ms.

A power electronic FLCB interrupts fault current by turning off insulated gate bipolar transistors (IGBTs), and uses a surge arrestor to absorb the voltage surge without forming an arc. There are no moving parts and no arc to interrupt, so the fault current can be interrupted within 2 ms or less.

However, most existing FLCBs suffer from limitations caused by conduction losses, as the IGBTs that interrupt fault current also have to carry normal load current. This means that they need many IGBT modules to handle the current at full load and/or need a large cooling system to dissipate heat at full load. This is why existing FLCBs are too large for use at typical London substations, and this characteristic is considered a showstopper.

ABB’s 2000 A FLCB concept eliminates conduction losses by using an innovative “fast commutating switch” (FCS) that bypasses the power electronics during normal operation, and opens within 0.35 ms in the event of a fault. It eliminates the need for a bulky cooling system, making the technology feasible to install in an existing indoor substation.

ABB proposes that the prototype can be housed in three 1 m wide modular switchgear cubicles. This is much smaller than other FLCB designs, with further size reductions possible for a commercial product.

The FCS also reduces network losses, which translates to lower operating costs.

The ABB FCS is of a novel design and has not yet been proven on a distribution network anywhere.

The ABB FLCB consists of the mechanical fast commutating switch, a power electronic switch and a surge arrester, all connected in parallel, as shown in Figure 2.

Power electronic switches such as IGBTs have very good switching properties and can turn off a current without waiting for a current zero crossing. However, power electronic switches or semiconductors are by definition poor conductors when compared with mechanical switches and therefore result in high losses when on. These losses are costly and also, as already noted, entail a large external cooling system to prevent the IGBTs from overheating. The mechanical switch on the other hand is not as good at switching and is not able to switch off a current prior to current zero crossing in order to achieve current limiting functionality. However, by combining the good properties of the power electronic switch and the mechanical switch, and excluding the poor properties, a very good FLCB is arrived at, with multi-shot capability.

The switch off sequence starts with the line current flowing through the closed mechanical switch to enable conduction of the nominal current with negligible losses. When a fault occurs and the FLCB is tripped, the mechanical switch is opened and the current is commutated into the power electronic switch, which is used for turning off the current in that branch. The current is then commutated to the parallel surge arrester and will produce a counter voltage based on the surge arrester protective level. This counter voltage will be higher than the system voltage and will force the current to zero before the natural current zero crossing and it will during that process also absorb the trapped magnetic energy in the system short circuit inductance.

The requirements on the fast mechanical commutating switch have until now not been fulfilled by any available switch. The switch needs to open very fast (<0.5 ms) since the short circuit current is rising fast. It should also be able to create a voltage high enough to ensure a fast and reliable commutation into the power electronic switches. The commutation voltage has to exceed the forward voltage drop of the power electronic switches and the voltage drop caused by the combination of the high di/dt during the commutation and the stray inductance of the loop formed by the switch and the power electronic switches. In addition, the switch in open position should withstand the voltage when the power electronic switches are turned off.

A suitable fast commutating switch has been developed by ABB, and is in fact part of the concept proposed for ABB’s “breakthrough” hybrid (power electronic/ mechanical) circuit breaker for HVDC.

The fast commutating switch utilises a novel contact system with a number of contacts connected in series. The number of series connected contacts ensure the high arc voltage required for the fast commutation. The series contacts also ensure a high electrical withstand with the short stroke of the contact system enabling fast operation. The contact system is connected to a bi- stable Thomson actuator which ensures both fast opening and fast closing.

AMAT’s 250A FLCB system currently forms part of its 2000 A solid-state fault current limiter, which uses a 250 A FLCB combined with a current-limiting mutual reactor to minimise physical size and conduction losses.

The 250 A FLCB will be trialled by itself (without the reactor), installed in front of a generator at a distributed generation site. It is thought this will be the first such UK installation at a generator’s premises (other than Is-limiters). Installing a FLCB at a generator’s premises completely avoids physical space constraints at existing substations. It also allows the customer to connect large amounts of generation even if the network is “full” because of fault level constraints.

Doing away with the reactor significantly reduces cost and physical size, but it does mean that the generator may be disconnected in the event of a network fault, and for some operators, this may be an unacceptable impact, Ofgem notes. For this reason, the project will also include investigation of whether generators would prefer a parallel/mutual reactor solution that enables a generator to “ride through” a network fault without contributing any significant fault current.